The steps that I am following are:

Step1: Ottawa Sand and 3% bentonite (by dry mass of sand) mixture is prepared by shaking the mixture in an airtight container.

Step2: The mixture is then air pluviated into a triaxial mold whose dimensions are height=10 cm and diameter=5 cm.

Step3: The sample is then confined on the triaxial machine.

Step4: The sample is flushed with CO2 at a small rate (3-4 bubbles a second) and then flushed with de-aired de-ionized water at a rate of 3 mL/min until 1 pore volume of water comes out of the sample.

Step5: The sample is kept still (under confinement) for 72 hours to ensure full hydration and swelling of bentonite.

- Before flushing the sample (when it is dry), bender element tests are performed. In this test, shear wave pulse is fired from the top piezoelectric element to the bottom piezoelectric element. Then, the time of travel of the wave is calculated by subtracting the time at which the first peak of the receiving oscillation appears from the time at which the transmitted peak is found. The calculated time of travel is used to calculate the shear wave velocity that is used to calculate the small strain shear modulus (Gmax) value.

The problem in my research is that after flushing the sample with water (5 minutes into flushing), no more clear received oscillations are found in order to find the time of travel. The same problem remains in the back saturation, consolidation, and shearing phases. However, before flushing the sample with water, I can read the received wave clearly.

One theory that I am trying to prove/disprove is the following:

- Since the bentonite absorbs a big amount of water, the shear waves can not travel from the top of the sample to the bottom of it because water can not support shear motions. However, what about the contact (interlocking forces) between the sand particles? Also, shouldn't the full hydration and swelling of bentonite form a gel-like material that can support shear waves?